1040 



Fishery Bulletin 97(4), 1999 



Only one species, C. pinnatibarbatus , is known to 

 have a broad, fimbriate membrane that unites the 

 barbels (Hubbs and Kampa, 1946; Imai, 1959). The 

 barbels become moderately long (Bbl > HL) in C. 

 atrisignis and C. spilonotopterus, whereas the other 

 species retain relatively short barbels (Bbl < HL). 

 The size and condition of the barbels in C. rapanouiensis 

 are unkpown, but in the closely related species C. agoo 

 they are short (Chen, 1987, 1988). 



Larval pigmentation in Cheilopogon initially 

 ranges from sparse to dense and typically increases 

 gradually during development. The sparsely pig- 

 mented larvae, including C. xenopterus, probably C. 

 atrisignis and C dorsomaculata, and possibly C. 

 spilonotopterus and C. papilio, display a pattern pri- 

 marily of melanophore rows on the dorsum, on the 

 horizontal septum, and on the ventral margin of the 

 tail, and usually bars form in the latter part of the 

 larval stage (Kovalevskaya, 1965, 1977; Chen, 1987. 

 1988). The more heavily pigmented species, includ- 

 ing C. furcatus, C. heterurus hubbsi, and C. 

 pinnatibarbatus californicus, display a pattern simi- 

 lar to that of larval Cypselurus, with melanophores 

 more evenly distributed over the trunk and at least 

 anteriorly on the tail (sometimes diminishing to dor- 

 sal, midlateral, and ventral rows on the posterior half 

 of the tail) (Hildebrand and Cable, 1930; Hubbs and 

 Kampa, 1946; Watson, 1996). Cheilopogon rapanou- 

 iensis might belong to this latter group as well, if its 

 larvae resemble the moderately pigmented larvae of 

 C. agoo (Chen, 1987, 1988). Juveniles of most spe- 

 cies are barred; about six bars (as in C. xenopterus) 

 is common. 



Larvae and juveniles of C. xenopterus usually can 

 be distinguished without great difficulty from the 

 other flyingfish species in the eastern Pacific, except 

 perhaps from C. dorsomaculata. The position of the 

 anal-fin origin below dorsal rays 4-7 distinguishes 

 C. xenopterus from Exocoetus, Fodiator, Hirun- 

 dichthys, and Oxyporhamphus, in which the anal- 

 fin origin ranges from just ahead of the dorsal-fin 

 origin to under dorsal rays 1-3, depending on spe- 

 cies (e.g. Imai, 1954; Kovalevskaya, 1964, 1980; Chen, 

 1988; Watson, 1996). Larger larvae and juveniles of 

 all four genera are further distinguished from C. 

 xenopterus (and all the other Cheilopogon species) 

 by their lack of paired mandibular barbels. Cheilo- 

 pogon xenopterus (and all the other flyingfishes) are 

 also distinguished from Exocoetus by having much 

 more posteriorly placed pelvic fins. Larval Fodiator 

 and Oxyporhamphus are unique in developing a beak 

 beginning at about 6 mm and 8 mm, respectively, 

 whereas Parexocoetus acquires a much smaller beak 

 in the juvenile stage, by about 18 mm (e.g. Collette 

 et al., 1984; Watson, 1996). Cheilopogon xenopterus 



is rather sparsely pigmented, in contrast to the more 

 general, denser pigmentation of larval Cypselurus, 

 Exocoetus, Fodiator, Parexocoetus, Prognichthys, and 

 the three or four Cheilopogon species noted above 

 (e.g., Hildebrand and Cable, 1930; Imai, 1959, 1960; 

 Kovalevskaya, 1980; Chen, 1988; Watson, 1996). 

 Among the more sparsely pigmented flyingfish lar- 

 vae, Hirundichthys (which may be moderately pig- 

 mented initially, but become more sparsely pig- 

 mented on at least the prepelvic part of the trunk; 

 e.g. Kovalevskaya, 1980) and Oxyporhamphus are 

 easily distinguished from C. xenopterus by the char- 

 acters noted above. In addition, larval Oxypor- 

 hamphus are more elongate, with much shorter 

 pectoral and pelvic fins, and they lack a row of mel- 

 anophores along the horizontal septum before about 

 9 mm (e.g. Khrapkova-Kovalevskaya, 1963; Watson, 

 1996) in contrast to before 4 mm in C. xenopterus. 

 Among the five Cheilopogon species in the eastern 

 Pacific with sparsely pigmented (or assumed to be 

 sparsely pigmented I larvae, it usually should be pos- 

 sible to distinguish C. atrisignis and C. papilio from 

 C. xenopterus by a combination of myomere and dor- 

 sal, anal, and (in larger larvae) pectoral-fin ray counts 

 (e.g. Table 1). At larger sizes (small larvae have not 

 been described) larval C. atrisignis and C spilono- 

 topterus can be distinguished from C. .xenopterus by 

 barbel structure and pigmentation, and C. papilio 

 can be distinguished by barbel structure and fin pig- 

 mentation, as noted above. Large larval and juve- 

 nile C. spilonotopterus are more fully and evenly pig- 

 mented on the trunk (by about 14 mm) and lack the 

 barred pattern displayed by larger lai-val and juve- 

 nile C. xenopterus (e.g. Kovalevskaya, 1977; Chen, 

 1987, 1988). Cheilopogon papilio likewise become 

 more generally pigmented than C. xenopterus, al- 

 though four or five bars remain visible through at 

 least 21 mm. Juvenile C. dorsomaculata smaller than 

 about 40 mm bear a striking resemblance to C 

 .xenopterus but can be distinguished by small differ- 

 ences in meristic characters (Table 1) and have more 

 sparsely pigmented (nearly unpigmented) pelvic fins 

 than do C. .xenopterus. Larvae of C. dorsomaculata have 

 not been described, but the similarity of the small ju- 

 veniles to C. .xenopterus suggests that it may be diffi- 

 cult to distinguish the larvae of the two species. 



Acknowledgments 



I thank the scientific and ship crews of the RV David 

 Starr Jordan and RV McArthur, and especially Bob 

 Pitman, Steve Reilly, and Tim Gerrodette of the 

 SWFSC Marine Mammals Division, for collecting and 

 providing access to the plankton samples containing 



